张宸轩

Key Skills

Software

No software listed

Top Subject Matter

E-commerce – Product Categorization & Customer Support

Logistics & Supply Chain Operations

Business Documents, Tables & Spreadsheet QA

Top Data Types

Document

Text

Top Task Types

Classification

Evaluation/Rating

Entity (NER) Classification

Freelancer Overview

Entry-level AI training and data QA candidate with related experience in structured data cleaning, customer-feedback classification, document/table review, spreadsheet QA, and evidence-based evaluation. Best-fit tasks include text classification, document QA, table review, entity extraction, evaluation/rating, and Chinese-English review. My background includes supply chain operations analysis, KPI dashboards, demand forecasting, inventory replenishment analysis, and high-volume customer feedback categorization. I am comfortable following detailed guidelines, checking consistency across records, organizing structured outputs, and writing concise English/Chinese summaries.

Labeling Experience

Demand Forecasting and Inventory Replenishment Strategy Optimization (Data Validation & Evaluation Metrics)

Validated forecasting outputs using quantitative evaluation metrics to support model comparison and selection. Transformed order-level details into a daily demand series and performed rolling backtesting to separate training and testing information. Documented inventory assumptions and scenario changes with clear quantitative rationale for evidence-based review. • Implemented forecasting method comparison with rolling backtesting for robust evaluation • Evaluated models with WMAPE, MAE, and RMSE and selected baselines by measured performance • Produced structured documentation of assumptions, safety stock, reorder points, and fill-rate changes • Demonstrated guideline-following and explainable evaluation practices relevant to rating tasks

2026 - 2026

Supply Chain Operations KPI Dashboard and Fulfillment Diagnosis (Data QA)

Text

Performed evidence-based evaluation by comparing rating outcomes between on-time and delayed orders to support QA-style judgment tasks. Cleaned e-commerce order datasets and organized KPI views to ensure consistent fields and category definitions for downstream review. Classified operational risk signals and delay patterns by region, product category, distance range, and month to generate follow-up review rules. • Used Excel and pivot tables to structure KPI views for validation and review • Converted operational observations into labeled follow-up review rules for high-risk segments • Applied evaluation thinking to compare outcomes across different delivery timing groups • Focused on guideline-consistent labeling logic and field consistency to support reliable outputs

2026 - 2026

Last-Mile Delivery and Fulfillment Operations Intern

DocumentDiagnosis

Observed last-mile delivery workflows and summarized fulfillment exceptions using consistent issue labels and resolution logic. Reviewed and documented cases such as address verification failures, customer absence, and appointment changes to support evidence-based QA. Applied accuracy and timeliness considerations when organizing exception records for downstream review. • Used standardized exception categories to describe delivery problems • Linked each exception type with resolution logic in structured notes • Summarized operational deviations with clear evidence similar to QA labeling workflows • Focused on consistent classification of real-world fulfillment issues

2024 - 2024

Customer Outreach and Operations Data Intern (Label-like customer feedback categorization)

TextClassification

Categorized high-volume customer call outcomes into intent, objections, and invalid contacts, aligning with text classification and feedback labeling patterns. Reviewed engagement trends and adjusted messaging sequences based on observed user reactions to improve categorization effectiveness. Maintained structured records of customer interactions for consistent evidence-based labeling. • Labeled customer-call outcomes using a fixed set of categories (intent/objections/invalid) • Analyzed response patterns to refine how categories were applied over time • Produced structured operational notes linking observations to labeled outcomes • Practiced consistent labeling logic for repeatable classification tasks

2024 - 2024

Education

Work History